Br. J. C(ancer (1976) 33, 400
DIAGNOSTIC AND PROGNOSTIC SIGNIFICANCE OF PERIPHERAL BLOOD CULTURAL CHARACTERISTICS IN ADULT ACUTE LEUKAEMIA F. H. BALKJrILL AND R. T. D. OLIV'ER
From the Imperial Cancer Research Fund Department of JMedical Oncology, St Bartholomew's Hospital, London E.C.1 Received 26 November 1975
Accepted 2 December 1975
Summary.-A simple liquid culture technique has been used to study peripheral blood from patients with acute myelogenous leukaemia. Evidence is presented that cells from morphologically identical types of leukaemia have differing capacity for " differentiation " from free floating blast cells into plastic-adherent, phagocytic, trypsin-resistant macrophage-like cells with Fc and C 3 receptors. Preliminary analysis suggests that patients whose cells have the greatest capacity for " differentiation " have a better chance of achieving complete remission.
UNCONTROLLED proliferation of primitive cells and failure of normal differentiation are thought to be basic defects leading to acute leukaemia. The use of in vitro culture is an obvious way to investigate this problem, but until the early 1 960s the majority of research workers found the culture of normal and abnormal haemopoietic cells to be very difficult (Woodliff, 1964). In a variety of systems (Woodliff, 1964; Woodliff, 1958; Woodliff, 1961; Osgood and Brooke, 1958; Lewis and Lewis, 1911; Carrel and Burrows, 1911) cells only survived a few days and showed little evidence of multiplication or differentiation. With the discovery of growth stimulating factors such as CSF (colony stimulating factor) and the introduction of a semisolid agar culture system, actual growth and differentiation of mouse (Pluznik and Sachs, 1965; Bradley and Metcalf, 1966) and human (Pike and Robinson, 1970) bone marrow cells were obtained. Despite some reports to the contrary (Senn, McCulloch and Till, 1967; Iscove et al., 1971) the agar system has been Reprint requests to: DI R. T. D.
Lomilon, E.Ci.
found to support the growth of acute and chronic granulocytic leukaemia cells (Paran et al., 1970; Moore, Williams and Metcalf, 1973; Robinson, Kurnik and Pike, 1971). Variation between reports is considerable (Senn et al., 1967; Paran et al., 1970; Moore et al., 1973; Robinson et al., 1971; Brown and Carbonne, 1971; Greenberg, Nichols and Shrier, 1971; Iscove et al., 1971), but this may be due to differences in method and the small number of patients in each study. In a study of 127 cases of pre-treatment acute leukaemias, Moore et al. (1974) reported a variety of growth patterns indicating a spectrum of normal to abnormal growth and differentiation. Although agar cultures have contributed much to our knowledge of normal and abnormal myelopoiesis and factors influencing it (Metcalfe, 1973), cellular mobility and interaction are not easily studied using this system and retrieval of cells is difficult. Therefore, it was a welcome advance when liquid culture systems for mouse (Sumner et al., 1972) and human bone marrow cells (Golde and Cline, 1973) which supported both growth Oliver, Department of AMedical Oncology, St. Bartholome-w's Hospital,
PERIPHERAL BLOOD CULTURAL CHARACTERISTICS IN LEUKAEMIA
and differentiation were reported. Using Marbrook diffusion chambers (Marbrook. 1967) Golde and Cline (1973) reported growth of normal human bone marrow and some leukaemic cells and showed that these cells were capable of growth in the absence of an external source of CSF. However, the complexity of the apparatus required limited the generalized use of this technique. Other workers have reported leukaemia cell growth in less complicated culture with, and in some cases, without stimulating factors (Aye, Till and McCulloch, 1972; Aye et al., 1974a; Aye, Till and McCulloch, 1974b), but again only small numbers of patients were studied and some of the patients had recently received chemotherapy. These workers found no evidence of cellular differentiation (Aye et al., 1974a). We reported in 1974 that in a simple quaintitative liquid microculture system the majority of AML cells would proliferate (Balkwill, Pindar and Crowther, 1974) and in this paper we wish to describe morphological and functional studies on these cells grown in both the m-icrosystem and larger culture vessels. AIATERIALS AND) METHODS
Preparation of the cells.-Fresh peripheral blasts from untreated patients with AML, or buffy-coat cells from normal individuals, were separated from h-eparinized blood (300 units mucous preservativ-e-free heparin to 10 ml blood) by the addition of 1% methyl cellulose (3 ml to 10 ml blood). This was allowed to sediment for 15 min at rooni temperature, when the buffy coat was removed and washed once in medium RPM1 1640 (Gribco Biocult Ltd). Liquid-nitrogenstored pre-treatment peripheral blasts were thawed at 37 °C and then diluted dropwise over 10 minutes with RPM1 1640 to ensure good viability as previously described (Balkwill et al., 1974; Williams, Ficat and Oliver, 1975). These cells were washed once. Culture of the cells.-For culture, cells wi-ere ie-suspended at 1 x 106/ml in medium 199 with antibiotics (Wellcome Reagents Ltd) supplemented by L-asparagine 0-2 27
401
mg/ml, extra L-glutamine (1 ml 200 nmM solution/100 ml medium) (Flow Laboratories Ltd) and 10% Foetal Bovine serum, Batch No. 417095 (Flow Laboratories Ltd). Ten ml aliquots of the cell suspen;icn were added to tissue culture flasks (Falccn Plastics 3013) or 0-2 ml aliquots added to Falcon 3040 microtitre plates. Sealed tissue culture flasks and plates with lids were incubated in 5°h CO2 85% humidity atmosphere at 37°C. Observation and Photomicroscopy.-Obser-
vation and photomicroscopy of living cells carried out under phase contrast microscopy using a Wild M40 inverted microscope equipped with a 35 mm camera attachment and using Agfapan 25 film. Fixed and stained cells were examined and photographed on either a Wild M20 microscope or a Leitz Ortholux 1 Microscope using Kodak photomicrograph 2483 film. Observcation of stained cells.-Non-adherent cells were smeared on glass slides or cytocentrifuge preparations made. Adherent monolayers were washed x 3 in Medium 199 and then fixed in methanol, stained in situ and then the sides and top of the bottle removed with pliers. The stained monolayers were mounted with Ceedol (Raymond Lamb) or neutral glycerin jelly (G. T. Gurr Ltd) as conventional slide mounts degraded the plastic. Assessment of the uptake of tritiated thymidine. -The cells grown in microculture were pulsed for 16 h with H3-thymidine sp. were
act. 5Ci/mmol (Amersham Radiochemicals) to give a final concentration in the wells of 0.5 ,uCi/ml. The cultures were harvested using a cell harvesting machine, the filter
papers dried and counted in an Inter-
technique ABAC SL40 Liquid Scintillation counter. Values obtained from 6 wells w-ere averaged for each estimation. Phagocytosis.-Cultures of Staph. albus
grown overnight in nutrient broth were washed x 2 in PBS and then resuspended in RPM1 1640 + 10% human AB serum. These were incubated at 37 °C for 30 min and then washed x 2 again before being added to the cells in the culture flasks. The cultures were incubated at 37°C for two hours and then the non-adherent cells + bacteria were removed from the bottles, washed x 3 with slow spinning in order to remove the majority of non-ingested bacteria, and then cytocentrifuge preparations were made of the cell pellet. Adherent cells were
402
F. R. BALKWILL AND R. T. D. OLIVER
washed x 3 in RPM1 1640 by gentle pipetting and then wet-fixed in methanol. All preparations were Gram stained and counterstained with neutral red. Trypsin resistance of adherent cells.-Nonadherent cells were removed from the cultures and the remaining adherent cells washed x 3 in RPM1 1640. The number of adherent cells in 8 marked fields on each bottle was then counted and Medium 199 containing 0.1% trypsin added for 10 min at 37°C. The monolayers were washed x 3 in Medium 199 and the remaining cells counted in the same fields. The percentage cell reduction due to the presence of trypsin was calculated. Receptor techniques.-E Sheep red blood cells (SRBC) preserved in Alsever's solution (Wellcome Reagents Ltd) were washed x 3 in medium before use. EA 1/100 dilution of a rabbit anti-sheep red cell serum (Wellcome Reagents Ltd) was added to an equal volume of 5% SRBC in Medium 199 and incubated 30 min at 370C. The coated SRBC were washed x 3 in Medium 199 and stored at 40C until used (not more than 4 h after making up). E-IgM and EAC Rabbit anti-SRBC IgM fraction was prepared by i.v. injection of SRBC washed x 2 in PBS. 0-5 ml of a 10% solution was injected and 7 days later serum collected. The euglobulin fraction of this serum was precipitated by ammonium sulphate (Heide and Schwick, 1973) and dialysed against Tris/HCl buffer pH 8-2 for one day. Column fractionation of a Sephadex G 200 column resulted in a single peak which was concentrated by an Amicon Diaflow Filter P30 (Amicon Ltd) and stored at -17900 until needed. A 1/100 dilution of this fraction in Medium 199 was added to an equal volume of a 5% SRBC suspension, E, and incubated 30 min at 370C. This E-IgM complex was washed x 3 in Medium 199. The source of complement was human AB serum stored at - 1790C in small aliquots. This was thawed immediately before use and equal volumes of a 1/100 dilution of this and the E-IgM were incubated at 370C for 15 min. Both the E-IgM and the complement were suspended in CFT buffer with 0-15 mM CaCl2 and 0-5 mM MgCl2. 6H20. After washing x 3 in Medium 199 the E-IgM and the EAC were stored at 40C until used (not more than 4 h after making up).
The microculture system was used for the evaluation of receptors, thus allowing only small volumes of reagents to be used and each test done in duplicate. The non-adherent cells were removed from the wells, washed x 2 and then 2 x 105 of these added to a 2% solution of E, E-IgM, EA(IgM)C in Beckman tubes. The Beckman tubes were then incubated at 370C with a rotation of 22 rev/min for 30 min and then allowed to settle. Rosettes were counted using a Wild M1O microscope under phase contrast and at least 400 cells on two different haemocytometer preps, counted for each parameter. After removal of the non-adherent cells, the adherent cells were washed x 3 by gentle pipetting in Medium 199 and then 2% solution of the above erythrocyte and antibody reagents added to each well. After 30 min incubation at 370C, unattached SRBC were removed by immersion of the plate several times into a bath of warm medium. Each parameter was assessed using a Wild M40 inverted microscope on duplicate wells with at least 400 cells being counted. All rosettes were scored as the attachment of five or more SRBC/cell. RESULTS
Microculture system.-Peripheral blood (PB) taken at the time of diagnosis from 48 out of 58 (83%) patients with AML grew in microculture with an incorporation from 1000 to 50,000 ct/min/well of 3H-thymidine measured after 4 days of culture. Macroculture-morphological observations.-In 12/60 (20%) successfully grown flask cultures from acute myelogenous leukaemia pretreatment peripheral blood, free-floating rounded cells persisted throughout the period of culture, but in the remaining 80% during the period of culture, elongated plastic-adherent cells developed (Fig. 1). The free-floating cells were found to survive up to 28 days in culture with frequent mitoses being observed, whereas the adherent cells survived up to 6 months, often forming multinucleate cells and showing no evidence of mitoses.
PERIPHERAL BLOOD CULTURAL CHARACTERISTICS IN LEUKAEMIA
'
:3
II
dIv.
403
I day
6
davs
FIG. 1.-Development of a leukaemia cell culture into adherent and non-adherent cells.
E;M~~~~~~~~~~~~~~~~~~~~~~-
.0k.
.......
x
150.
....SEiL
FiG. 2.-May-Griinwald-Giemsa stained cytocentrifuge preparations of supernatant populations from leukaemic peripheral blood after 7 days culture showing mainly undifferentiated cells. x 135.
404
F. R. BALKWILL AND R. T. D. OLIVER
May-Grtinwald-Giemsa (MGG) stained preparations of the non-adherent cells revealed a large variety and spectrum of normal and abnormal differentiation of monocyte and granulocytes, precursor cells (Fig. 2) and in some cultures a, few mature polymorphs (Fig. 3). The adherent population, however, had the appearance of mature mononuclear phagocytes (see Fig. 1). In order to investigate the relevance of these two populations of cells to the lisease, each pre-treatment fresh PB culture was classified as to the number of typical adherent cells per x ]50 high power phase contrast field. The cultures were always scored on the 3rd or 4th day of growth. The cultures were classified as shown in Fig. 4. + + + cultures >50, + + cultures 20-50, + cultures 1-20, - cultures < 1 fusiform cell per high power field.
Functional studies. The liquidl nitrogen stored cells when cultured showed the same characteristic morphology as cells that had not been frozen, and these have been used for the detailed functional studies described below. All cell preparations had greater than 9500 viability assessed by trypani blue dye exclusion. Phagocytosis. The adherent cells wer e found to be avidly phagocytic for the opsonized Staph. albus particles (Fig. 5) whereas the majority of uioin-adherecnt cells were poorly phagocytic. Six leuikaemias and one normal were studied aiid the results shown in Table I. Trypsin resistance. The results in Table II indicate the peticentage of plastic-adherent cells which are trypsinresistant in + + + and + cultures an(l( compares this with normal peripheral blood cultures and human embrvo fibroblasts. It is clear that in the +
::....
.;I
'- ']F ::9',^!. :: : ....~~~.......
,. ... . ... : =, ........................... .W.;0;td: .......
......
;\0;4;;itE
....:
0;fltt
. . . . . . . .... . . . .... iVliD;'ttt0\00011Xi;fEtdDtEji\iV000.
..............................................._...
FIG. 3.-May-Griinwald-Giemsa stained cytocentriftige preparations of supernatant from letukaemic peripheral blood after 7 days cuiltuire showinig some myeloi(l alIl diffi(rlntiation. x 1 35.
populatioI1s Ifl)monocytoil
P
ER{IPHERAL
Fic(. 4.-
Ic.
5.
BLOOI)
405
CULTURAL CHARACTERISTICS IN LEUKAEMIA
ifferent types of letikaemia peripheral blood cultures.
x
150.
Phalgocytosis of Staph. (dlbeis by a(dheieiit cells from 4-day leukaemia cell culture.
x
135.
406
F. R. BALKWILL AND R. T. D. OLIVER
TABLE I.-Phayocytosis Experiments on Cells from Patients with AML Day 1 Patient no. 1 2 3 4 5 6
Classification of cell culture +++ +++ +++ +++ + _ Normal control
Adherent cells 57* 95
Supernatant cells 63 24 NT NT 3 4 32
44 40 5 No cells 83
Day 4 Adherent Supernatant cells 80 97 100 98 48 No cells 98
cells 36 18 19 32 16 3 NT
NT. Not tested. * Percentage of cells containing ingested Staph. albus. For explanation of classification see Fig. 4.
TABLE II.-Trypsin Resistance Experiment on Cells from Patients with AML Percentage trypsin-resistant adherent cells 89 79 88 84 +-++ 44 + 38 + 13 + Normal control 87 Human embryo fibroblasts Less than 1 For explanation of classification see Fig. 4.
Patient no. 1 2 3 4 5 6 7
Classification of cell culture +++ +++ +++
cultures, plastic-adherent cells are as resistant to trypsin as normal peripheral blood macrophages, while the majority of plastic-adherent cells in + cultures are removed by trypsin treatment. Receptors.-Eleven liquid nitrogen stored blast cell cultures (3 + + +, 6 +, 2 -) and 3 fresh normal cell cultures were set up as described in Methods to assess the presence of the Fc and C3 receptor on these cells pre-culture, and Days 1, 4-5 and 7-8 of culture. After culture a high proportion of the adherent cells were found to have both C3 and Fc receptors as Fig. 6 shows. The finding that the EA rosettes were invariably phagocytosed during the incubation time, whereas the EAC rosettes were never ingested, was unequivocal throughout the whole series of experiments. Also, the value of E or E-IgM rosettes was never greater than 1%.
A summary of the results obtained in the series of receptor experiments is shown in Fig. 7. These results clearly show that mature cells possessing the C3 and Fc receptor first developed in the supernatants of the + + + cultures, and then increasing numbers were found amongst the plastic-adherent cells. No differentiation of this kind was found in the +, - or the normal cultures studied, although at 7 days a few receptor positive cells did appear in the floating population. Relationship of type of culture to diagnosis.-Table III shows the relationship of type of culture to diagnosis and indicates that although there is a definite tendency towards a greater percentage of adherent cells in patients with acute myelomonocytic leukaemia, this does not always occur. In addition, many of the so-called acute myelogenous leukaemias (32%) have quite large proportions of adherent cells (the + + + and + + groups). Cultural characteristics and response to treatment.-The patients in this study received standard chemotherapy induction using protocols which have been reported in detail elsewhere (Crowther et al., 1973). Table IV shows the incidence of complete remission achieved in these patients relative to the cultural type. Fifty-four per cent of patients in the +A+A+ and +A+ group, 18% of the + and - group and 15% in the no growth
407
PERIPHERAL BLOOD CULTURAL CHARACTERISTICS IN LEUKAEMIA
'1',.A
.
A(
FIG. 6.-The demonstration of Fc and C3 receptors on adherent cells from 7-day leukaemia cell cultures. x 150.
TABLE III.-Relationship of Cultural Cha- group achieved racteristics to Haematological Diagnosis (P < 0*025). Stickers ++ Stickers +
1
9
15
6
13
4
11
1
10
0
Stickers
Non-stickers No growth
Myeloblastic Myelomonocytic For explanation of classification see Fig. 4.
complete
remission
DISCUSSION
The studies we have so far carried out on the cultural characteristics of acute myelogenous leukaemia cells in the peripheral blood have confirmed, firstly that these cells will proliferate actively in vitro without any external growth stimulating factors, as has been
408
F. R. BALKWILL AND R. T. D. OLIVER
LEUKAEMIC
PERIPHERAL BLOOD NORMAL PERIPHERAL BLOOD
-------
- *- - E A 0
-
ROSETTE
EAC
ROSETTE
NON STICKERS (2)
I2
TIME IN DAYS
3
4
5
6
7
8
L Pr
FIG. 7.-Evolution of receptors in cell cultures.
TABLE IV Cultural characteristics
+++
+ and
and
No. of
patients
++
Achievement of complete remission
24 27 7
-
No growth P
DIAGNOSTIC AND PROGNOSTIC SIGNIFICANCE OF PERIPHERAL BLOOD CULTURAL CHARACTERISTICS IN ADULT ACUTE LEUKAEMIA F. H. BALKJrILL AND R. T. D. OLIV'ER
From the Imperial Cancer Research Fund Department of JMedical Oncology, St Bartholomew's Hospital, London E.C.1 Received 26 November 1975
Accepted 2 December 1975
Summary.-A simple liquid culture technique has been used to study peripheral blood from patients with acute myelogenous leukaemia. Evidence is presented that cells from morphologically identical types of leukaemia have differing capacity for " differentiation " from free floating blast cells into plastic-adherent, phagocytic, trypsin-resistant macrophage-like cells with Fc and C 3 receptors. Preliminary analysis suggests that patients whose cells have the greatest capacity for " differentiation " have a better chance of achieving complete remission.
UNCONTROLLED proliferation of primitive cells and failure of normal differentiation are thought to be basic defects leading to acute leukaemia. The use of in vitro culture is an obvious way to investigate this problem, but until the early 1 960s the majority of research workers found the culture of normal and abnormal haemopoietic cells to be very difficult (Woodliff, 1964). In a variety of systems (Woodliff, 1964; Woodliff, 1958; Woodliff, 1961; Osgood and Brooke, 1958; Lewis and Lewis, 1911; Carrel and Burrows, 1911) cells only survived a few days and showed little evidence of multiplication or differentiation. With the discovery of growth stimulating factors such as CSF (colony stimulating factor) and the introduction of a semisolid agar culture system, actual growth and differentiation of mouse (Pluznik and Sachs, 1965; Bradley and Metcalf, 1966) and human (Pike and Robinson, 1970) bone marrow cells were obtained. Despite some reports to the contrary (Senn, McCulloch and Till, 1967; Iscove et al., 1971) the agar system has been Reprint requests to: DI R. T. D.
Lomilon, E.Ci.
found to support the growth of acute and chronic granulocytic leukaemia cells (Paran et al., 1970; Moore, Williams and Metcalf, 1973; Robinson, Kurnik and Pike, 1971). Variation between reports is considerable (Senn et al., 1967; Paran et al., 1970; Moore et al., 1973; Robinson et al., 1971; Brown and Carbonne, 1971; Greenberg, Nichols and Shrier, 1971; Iscove et al., 1971), but this may be due to differences in method and the small number of patients in each study. In a study of 127 cases of pre-treatment acute leukaemias, Moore et al. (1974) reported a variety of growth patterns indicating a spectrum of normal to abnormal growth and differentiation. Although agar cultures have contributed much to our knowledge of normal and abnormal myelopoiesis and factors influencing it (Metcalfe, 1973), cellular mobility and interaction are not easily studied using this system and retrieval of cells is difficult. Therefore, it was a welcome advance when liquid culture systems for mouse (Sumner et al., 1972) and human bone marrow cells (Golde and Cline, 1973) which supported both growth Oliver, Department of AMedical Oncology, St. Bartholome-w's Hospital,
PERIPHERAL BLOOD CULTURAL CHARACTERISTICS IN LEUKAEMIA
and differentiation were reported. Using Marbrook diffusion chambers (Marbrook. 1967) Golde and Cline (1973) reported growth of normal human bone marrow and some leukaemic cells and showed that these cells were capable of growth in the absence of an external source of CSF. However, the complexity of the apparatus required limited the generalized use of this technique. Other workers have reported leukaemia cell growth in less complicated culture with, and in some cases, without stimulating factors (Aye, Till and McCulloch, 1972; Aye et al., 1974a; Aye, Till and McCulloch, 1974b), but again only small numbers of patients were studied and some of the patients had recently received chemotherapy. These workers found no evidence of cellular differentiation (Aye et al., 1974a). We reported in 1974 that in a simple quaintitative liquid microculture system the majority of AML cells would proliferate (Balkwill, Pindar and Crowther, 1974) and in this paper we wish to describe morphological and functional studies on these cells grown in both the m-icrosystem and larger culture vessels. AIATERIALS AND) METHODS
Preparation of the cells.-Fresh peripheral blasts from untreated patients with AML, or buffy-coat cells from normal individuals, were separated from h-eparinized blood (300 units mucous preservativ-e-free heparin to 10 ml blood) by the addition of 1% methyl cellulose (3 ml to 10 ml blood). This was allowed to sediment for 15 min at rooni temperature, when the buffy coat was removed and washed once in medium RPM1 1640 (Gribco Biocult Ltd). Liquid-nitrogenstored pre-treatment peripheral blasts were thawed at 37 °C and then diluted dropwise over 10 minutes with RPM1 1640 to ensure good viability as previously described (Balkwill et al., 1974; Williams, Ficat and Oliver, 1975). These cells were washed once. Culture of the cells.-For culture, cells wi-ere ie-suspended at 1 x 106/ml in medium 199 with antibiotics (Wellcome Reagents Ltd) supplemented by L-asparagine 0-2 27
401
mg/ml, extra L-glutamine (1 ml 200 nmM solution/100 ml medium) (Flow Laboratories Ltd) and 10% Foetal Bovine serum, Batch No. 417095 (Flow Laboratories Ltd). Ten ml aliquots of the cell suspen;icn were added to tissue culture flasks (Falccn Plastics 3013) or 0-2 ml aliquots added to Falcon 3040 microtitre plates. Sealed tissue culture flasks and plates with lids were incubated in 5°h CO2 85% humidity atmosphere at 37°C. Observation and Photomicroscopy.-Obser-
vation and photomicroscopy of living cells carried out under phase contrast microscopy using a Wild M40 inverted microscope equipped with a 35 mm camera attachment and using Agfapan 25 film. Fixed and stained cells were examined and photographed on either a Wild M20 microscope or a Leitz Ortholux 1 Microscope using Kodak photomicrograph 2483 film. Observcation of stained cells.-Non-adherent cells were smeared on glass slides or cytocentrifuge preparations made. Adherent monolayers were washed x 3 in Medium 199 and then fixed in methanol, stained in situ and then the sides and top of the bottle removed with pliers. The stained monolayers were mounted with Ceedol (Raymond Lamb) or neutral glycerin jelly (G. T. Gurr Ltd) as conventional slide mounts degraded the plastic. Assessment of the uptake of tritiated thymidine. -The cells grown in microculture were pulsed for 16 h with H3-thymidine sp. were
act. 5Ci/mmol (Amersham Radiochemicals) to give a final concentration in the wells of 0.5 ,uCi/ml. The cultures were harvested using a cell harvesting machine, the filter
papers dried and counted in an Inter-
technique ABAC SL40 Liquid Scintillation counter. Values obtained from 6 wells w-ere averaged for each estimation. Phagocytosis.-Cultures of Staph. albus
grown overnight in nutrient broth were washed x 2 in PBS and then resuspended in RPM1 1640 + 10% human AB serum. These were incubated at 37 °C for 30 min and then washed x 2 again before being added to the cells in the culture flasks. The cultures were incubated at 37°C for two hours and then the non-adherent cells + bacteria were removed from the bottles, washed x 3 with slow spinning in order to remove the majority of non-ingested bacteria, and then cytocentrifuge preparations were made of the cell pellet. Adherent cells were
402
F. R. BALKWILL AND R. T. D. OLIVER
washed x 3 in RPM1 1640 by gentle pipetting and then wet-fixed in methanol. All preparations were Gram stained and counterstained with neutral red. Trypsin resistance of adherent cells.-Nonadherent cells were removed from the cultures and the remaining adherent cells washed x 3 in RPM1 1640. The number of adherent cells in 8 marked fields on each bottle was then counted and Medium 199 containing 0.1% trypsin added for 10 min at 37°C. The monolayers were washed x 3 in Medium 199 and the remaining cells counted in the same fields. The percentage cell reduction due to the presence of trypsin was calculated. Receptor techniques.-E Sheep red blood cells (SRBC) preserved in Alsever's solution (Wellcome Reagents Ltd) were washed x 3 in medium before use. EA 1/100 dilution of a rabbit anti-sheep red cell serum (Wellcome Reagents Ltd) was added to an equal volume of 5% SRBC in Medium 199 and incubated 30 min at 370C. The coated SRBC were washed x 3 in Medium 199 and stored at 40C until used (not more than 4 h after making up). E-IgM and EAC Rabbit anti-SRBC IgM fraction was prepared by i.v. injection of SRBC washed x 2 in PBS. 0-5 ml of a 10% solution was injected and 7 days later serum collected. The euglobulin fraction of this serum was precipitated by ammonium sulphate (Heide and Schwick, 1973) and dialysed against Tris/HCl buffer pH 8-2 for one day. Column fractionation of a Sephadex G 200 column resulted in a single peak which was concentrated by an Amicon Diaflow Filter P30 (Amicon Ltd) and stored at -17900 until needed. A 1/100 dilution of this fraction in Medium 199 was added to an equal volume of a 5% SRBC suspension, E, and incubated 30 min at 370C. This E-IgM complex was washed x 3 in Medium 199. The source of complement was human AB serum stored at - 1790C in small aliquots. This was thawed immediately before use and equal volumes of a 1/100 dilution of this and the E-IgM were incubated at 370C for 15 min. Both the E-IgM and the complement were suspended in CFT buffer with 0-15 mM CaCl2 and 0-5 mM MgCl2. 6H20. After washing x 3 in Medium 199 the E-IgM and the EAC were stored at 40C until used (not more than 4 h after making up).
The microculture system was used for the evaluation of receptors, thus allowing only small volumes of reagents to be used and each test done in duplicate. The non-adherent cells were removed from the wells, washed x 2 and then 2 x 105 of these added to a 2% solution of E, E-IgM, EA(IgM)C in Beckman tubes. The Beckman tubes were then incubated at 370C with a rotation of 22 rev/min for 30 min and then allowed to settle. Rosettes were counted using a Wild M1O microscope under phase contrast and at least 400 cells on two different haemocytometer preps, counted for each parameter. After removal of the non-adherent cells, the adherent cells were washed x 3 by gentle pipetting in Medium 199 and then 2% solution of the above erythrocyte and antibody reagents added to each well. After 30 min incubation at 370C, unattached SRBC were removed by immersion of the plate several times into a bath of warm medium. Each parameter was assessed using a Wild M40 inverted microscope on duplicate wells with at least 400 cells being counted. All rosettes were scored as the attachment of five or more SRBC/cell. RESULTS
Microculture system.-Peripheral blood (PB) taken at the time of diagnosis from 48 out of 58 (83%) patients with AML grew in microculture with an incorporation from 1000 to 50,000 ct/min/well of 3H-thymidine measured after 4 days of culture. Macroculture-morphological observations.-In 12/60 (20%) successfully grown flask cultures from acute myelogenous leukaemia pretreatment peripheral blood, free-floating rounded cells persisted throughout the period of culture, but in the remaining 80% during the period of culture, elongated plastic-adherent cells developed (Fig. 1). The free-floating cells were found to survive up to 28 days in culture with frequent mitoses being observed, whereas the adherent cells survived up to 6 months, often forming multinucleate cells and showing no evidence of mitoses.
PERIPHERAL BLOOD CULTURAL CHARACTERISTICS IN LEUKAEMIA
'
:3
II
dIv.
403
I day
6
davs
FIG. 1.-Development of a leukaemia cell culture into adherent and non-adherent cells.
E;M~~~~~~~~~~~~~~~~~~~~~~-
.0k.
.......
x
150.
....SEiL
FiG. 2.-May-Griinwald-Giemsa stained cytocentrifuge preparations of supernatant populations from leukaemic peripheral blood after 7 days culture showing mainly undifferentiated cells. x 135.
404
F. R. BALKWILL AND R. T. D. OLIVER
May-Grtinwald-Giemsa (MGG) stained preparations of the non-adherent cells revealed a large variety and spectrum of normal and abnormal differentiation of monocyte and granulocytes, precursor cells (Fig. 2) and in some cultures a, few mature polymorphs (Fig. 3). The adherent population, however, had the appearance of mature mononuclear phagocytes (see Fig. 1). In order to investigate the relevance of these two populations of cells to the lisease, each pre-treatment fresh PB culture was classified as to the number of typical adherent cells per x ]50 high power phase contrast field. The cultures were always scored on the 3rd or 4th day of growth. The cultures were classified as shown in Fig. 4. + + + cultures >50, + + cultures 20-50, + cultures 1-20, - cultures < 1 fusiform cell per high power field.
Functional studies. The liquidl nitrogen stored cells when cultured showed the same characteristic morphology as cells that had not been frozen, and these have been used for the detailed functional studies described below. All cell preparations had greater than 9500 viability assessed by trypani blue dye exclusion. Phagocytosis. The adherent cells wer e found to be avidly phagocytic for the opsonized Staph. albus particles (Fig. 5) whereas the majority of uioin-adherecnt cells were poorly phagocytic. Six leuikaemias and one normal were studied aiid the results shown in Table I. Trypsin resistance. The results in Table II indicate the peticentage of plastic-adherent cells which are trypsinresistant in + + + and + cultures an(l( compares this with normal peripheral blood cultures and human embrvo fibroblasts. It is clear that in the +
::....
.;I
'- ']F ::9',^!. :: : ....~~~.......
,. ... . ... : =, ........................... .W.;0;td: .......
......
;\0;4;;itE
....:
0;fltt
. . . . . . . .... . . . .... iVliD;'ttt0\00011Xi;fEtdDtEji\iV000.
..............................................._...
FIG. 3.-May-Griinwald-Giemsa stained cytocentriftige preparations of supernatant from letukaemic peripheral blood after 7 days cuiltuire showinig some myeloi(l alIl diffi(rlntiation. x 1 35.
populatioI1s Ifl)monocytoil
P
ER{IPHERAL
Fic(. 4.-
Ic.
5.
BLOOI)
405
CULTURAL CHARACTERISTICS IN LEUKAEMIA
ifferent types of letikaemia peripheral blood cultures.
x
150.
Phalgocytosis of Staph. (dlbeis by a(dheieiit cells from 4-day leukaemia cell culture.
x
135.
406
F. R. BALKWILL AND R. T. D. OLIVER
TABLE I.-Phayocytosis Experiments on Cells from Patients with AML Day 1 Patient no. 1 2 3 4 5 6
Classification of cell culture +++ +++ +++ +++ + _ Normal control
Adherent cells 57* 95
Supernatant cells 63 24 NT NT 3 4 32
44 40 5 No cells 83
Day 4 Adherent Supernatant cells 80 97 100 98 48 No cells 98
cells 36 18 19 32 16 3 NT
NT. Not tested. * Percentage of cells containing ingested Staph. albus. For explanation of classification see Fig. 4.
TABLE II.-Trypsin Resistance Experiment on Cells from Patients with AML Percentage trypsin-resistant adherent cells 89 79 88 84 +-++ 44 + 38 + 13 + Normal control 87 Human embryo fibroblasts Less than 1 For explanation of classification see Fig. 4.
Patient no. 1 2 3 4 5 6 7
Classification of cell culture +++ +++ +++
cultures, plastic-adherent cells are as resistant to trypsin as normal peripheral blood macrophages, while the majority of plastic-adherent cells in + cultures are removed by trypsin treatment. Receptors.-Eleven liquid nitrogen stored blast cell cultures (3 + + +, 6 +, 2 -) and 3 fresh normal cell cultures were set up as described in Methods to assess the presence of the Fc and C3 receptor on these cells pre-culture, and Days 1, 4-5 and 7-8 of culture. After culture a high proportion of the adherent cells were found to have both C3 and Fc receptors as Fig. 6 shows. The finding that the EA rosettes were invariably phagocytosed during the incubation time, whereas the EAC rosettes were never ingested, was unequivocal throughout the whole series of experiments. Also, the value of E or E-IgM rosettes was never greater than 1%.
A summary of the results obtained in the series of receptor experiments is shown in Fig. 7. These results clearly show that mature cells possessing the C3 and Fc receptor first developed in the supernatants of the + + + cultures, and then increasing numbers were found amongst the plastic-adherent cells. No differentiation of this kind was found in the +, - or the normal cultures studied, although at 7 days a few receptor positive cells did appear in the floating population. Relationship of type of culture to diagnosis.-Table III shows the relationship of type of culture to diagnosis and indicates that although there is a definite tendency towards a greater percentage of adherent cells in patients with acute myelomonocytic leukaemia, this does not always occur. In addition, many of the so-called acute myelogenous leukaemias (32%) have quite large proportions of adherent cells (the + + + and + + groups). Cultural characteristics and response to treatment.-The patients in this study received standard chemotherapy induction using protocols which have been reported in detail elsewhere (Crowther et al., 1973). Table IV shows the incidence of complete remission achieved in these patients relative to the cultural type. Fifty-four per cent of patients in the +A+A+ and +A+ group, 18% of the + and - group and 15% in the no growth
407
PERIPHERAL BLOOD CULTURAL CHARACTERISTICS IN LEUKAEMIA
'1',.A
.
A(
FIG. 6.-The demonstration of Fc and C3 receptors on adherent cells from 7-day leukaemia cell cultures. x 150.
TABLE III.-Relationship of Cultural Cha- group achieved racteristics to Haematological Diagnosis (P < 0*025). Stickers ++ Stickers +
1
9
15
6
13
4
11
1
10
0
Stickers
Non-stickers No growth
Myeloblastic Myelomonocytic For explanation of classification see Fig. 4.
complete
remission
DISCUSSION
The studies we have so far carried out on the cultural characteristics of acute myelogenous leukaemia cells in the peripheral blood have confirmed, firstly that these cells will proliferate actively in vitro without any external growth stimulating factors, as has been
408
F. R. BALKWILL AND R. T. D. OLIVER
LEUKAEMIC
PERIPHERAL BLOOD NORMAL PERIPHERAL BLOOD
-------
- *- - E A 0
-
ROSETTE
EAC
ROSETTE
NON STICKERS (2)
I2
TIME IN DAYS
3
4
5
6
7
8
L Pr
FIG. 7.-Evolution of receptors in cell cultures.
TABLE IV Cultural characteristics
+++
+ and
and
No. of
patients
++
Achievement of complete remission
24 27 7
-
No growth P
